Keywords
neuroendocrine neoplasms - insulinoma-associated protein 1 - immunohistochemistry - traditional neuroendocrine markers - H-score
Introduction
Neuroendocrine neoplasms (NENs) are heterogeneous group of disorders with varied histological patterns and nomenclature.[1] Recently, the incidence of NENs has been increased from an estimated 1.09 per 100,000 people in 1973 to 6.98 per 100,000 people in 2012 in the United States.[2] According to the Surveillance, Epidemiology, and End Results[3] and Indian[4] data more than 60% of neuroendocrine tumors (NETs) arise from the gastroenteropancreatic NETs (GEP-NETs). Clinical course of NENs is different and depends upon the location of the tumor; however, a significant number of patients can present with advanced stage.[5] All NENs share a common neuroendocrine origin and have varied organ-specific characteristics, biological behavior, prognosis, and treatment.[5] Diagnosis of NENs requires an integrated approach of pathological, immunohistochemical, genetic, and molecular markers.[5]
Insulinoma-associated protein 1 (INSM1) is a zinc-finger transcription factor which has a key role in the development of neuroendocrine differentiation in various tissues.[6] Insulinoma-associated-1gene encodes the INSM1, which was first discovered in 1992 at the National Institutes of Health (Bethesda, Maryland, United States) in human pancreatic insulinoma tissue and murine insulinoma cell lines.[7] Rosenbaum et al[8] reported, INSM1 as a robust immunohistochemical marker of neuroendocrine differentiation in normal and neoplastic human tissue. INSM1 is the first and the most widely validated pan-neuroendocrine marker which shows nuclear positivity.
There is a paucity of Indian literature on this new and emerging marker. In the present study, we will investigate the expression of INSM1 in NENs and compare it with the already established neuroendocrine markers.
Materials and Methods
Sample size and study design: This was a retrospective cross-sectional study done at the department of oncopathology in a tertiary cancer center. Consecutive 100 cases of confirmed NENs from November 2019 to January 2021 were enrolled in the study.
Inclusion and exclusion criteria: Immunohistochemically proven cases of NENs were included in the study. Tumors showing neuroendocrine differentiation without immunohistochemical confirmation and inadequate tissue, suspicious lesions, and cytologically diagnosed cases were excluded.
Demographic details were retrieved from the hospital database. All tissues were fixed in 10% buffered formalin and processed for hematoxylin and eosin and immunohistochemical study. NENs are classified according to the World Health Organization classification.[9] Immunostaining using synaptophysin (SYN) (SP11, monoclonal antibody, Thermoscientific, 1:50), chromogranin A (CgA) (LK2H10, monoclonal antibody, Thermoscientific, 1:100), and INSM1 (clone: MRQ-70, rabbit monoclonal antibody, Cell Marque, 1:50) antibodies were done on all cases. Nuclear immunoreactivity for INSM1 and cytoplasmic stain for SYN and CgA in tumor cells were considered positive. For all markers, both the percentage of cells and intensity of immunoreactivity were noted. H-score assessment was done for INSM1, SYN, and CgA.[10]
[11]
[12]
Statistical Analysis
Age, sex, location of tumor, histologic type, and histological grade were noted. Associations between categorical variables (location of tumor, tumor subtype, tumor grade) were analyzed using chi-square test. Wilcoxon rank test was used for comparison of H-score value. Two-sided p-values of < 0.05 were considered significant. All statistical analyses were carried out using SPSS 20.
Ethics: The institutional review committee of the Gujarat Cancer and Research Institute approved the study, approval number IRC/35/2019 and date November 14, 2019. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
Results
Clinicohistopathological and immunohistochemical characteristics are described in [Table 1]. The mean (±standard deviation [SD]) age was 55.5 (±10.61) years ranging from 25 to 76 years with most of the cases between 51 and 60 years (37%). Male preponderance was noted with male-to-female ratio of 4.2:1. Lung was the most frequently affected organ (59%). Total 67% cases had poorly differentiated neuroendocrine carcinoma (PD-NEC) of which 65% had small cell morphology. Total 43% of cases had a maximum tumor size of > 7 cm. Total 97 cases were positive for INSM1. Of three negative cases, two were PD-NEC of the lung and one was jejunum well-differentiated NET. Of these three negative INSM1 cases, one case was positive for both SYN and CgA, while two were positive of SYN and CgA, respectively.
Table 1
Clinicopathological and immunohistochemical characteristics
|
Parameters
|
Number of cases (n = 100)
|
|
Age
|
|
|
< 50 y
|
33
|
|
≥ 50 y
|
67
|
|
Site
|
|
|
Lung
|
59
|
|
GIT + pancreas
|
31
|
|
Others
|
10
|
|
Sex
|
|
|
Male
|
81
|
|
Female
|
19
|
|
Types
|
|
|
NET
|
29
|
|
NEC
|
67
|
|
Combined
|
4
|
|
Subtypes
|
|
|
NET
|
|
|
NET 1
|
16
|
|
NET2
|
5
|
|
NET3
|
8
|
|
MINEN
|
1
|
|
NEC
|
|
|
SMCC
|
65
|
|
LCNEC
|
2
|
|
Combined
|
3
|
|
Immunohistochemistry
|
|
|
Synaptophysin
|
|
|
Positive
|
96
|
|
Negative
|
04
|
|
Chromogranin
|
|
|
Positive
|
86
|
|
Negative
|
14
|
|
INSM1
|
|
|
Positive
|
97
|
|
Negative
|
03
|
Abbreviations: GIT, gastrointestinal tract; INSM1, insulinoma-associated protein 1; LCNEC, large cell neuroendocrine carcinoma; MINEN, mixed neuroendocrine nonneuroendocrine neoplasms; NEC, neuroendocrine carcinoma; NET, neuroendocrine tumor; SMCC, small cell carcinoma.
Correlation of Expression of INSM1 with Tumor Characteristics
Correlation of Expression of INSM1 with Tumor Characteristics
Association between various variables is described in [Table 2]. INSM1 expression was compared with size, site, histological type, histological grade, histological subtypes, and immunohistochemical markers. Statistically significant association was found between the expression of INSM1 and SYN and CgA (p-value < 0.05). However, no statistically significant association was found between histological type, histological grade, and histological subtypes (p-value > 0.05). Comparison of INSM1 with traditional NE markers is shown in [Supp. Figs. S1] and [S2].
Table 2
Correlation of INSM1 with clinicopathological variable
|
Variables
|
Total cases
|
Number of cases (%)
|
p-Value[a]
|
|
|
INSM1
positive
|
INSM1 negative
|
|
|
Histological types
|
|
|
|
|
|
NET
NEC
Combined
|
29
|
28 (96.6)
|
1 (3.4)
|
00591
|
|
67
|
65 (97.0)
|
2 (3.0)
|
|
4
|
4 (100)
|
0 (0)
|
|
Site
|
|
|
|
|
|
Lung
GIT and pancreas
Others
|
59
|
57 (96.6)
|
2 (3.4)
|
0.841
|
|
31
|
30 (96.8)
|
1 (3.2)
|
|
10
|
10 (100)
|
0 (0)
|
|
SYN
|
|
|
|
|
|
Positive
Negative
|
96
|
94 (97.9)
|
2 (2.1)
|
0.008
|
|
4
|
3 (75)
|
1 (25)
|
|
CgA
|
|
|
|
|
|
Positive
Negative
|
86
|
83 (96.5)
|
3 (3.5)
|
0.011
|
|
14
|
14 (100)
|
0 (0)
|
Abbreviations: CgA, chromogranin A; GIT, gastrointestinal tract; INSM1, insulinoma-associated protein 1; NEC, neuroendocrine carcinoma; NET, neuroendocrine tumor; SYN, synaptophysin.
a Chi-square test.
Comparisons of H-Scores
Detailed site-specific H-score calculation for all three antibodies is discussed in [Table 3]. Minimum and maximum H-score for all three antibodies was 0 and 300, respectively. Mean (±SD) H-score of SYN, CgA, and INSM1 was 119.65 (±78.706), 108.70 (±99.307), and 194.55 (77.818), respectively. Comparison of H-score is described in [Table 4].
Table 3
H-score comparison of SYN, CgA, and INSM1 by location of tumor
|
Tumor type
|
Antibodies
|
Positive cases
|
H-score
(mean)
|
|
GIT
|
SYN
|
21/21
|
135
|
|
CHR
|
18/21
|
128
|
|
INSM1
|
21/21
|
188
|
|
Pancreaticobiliary system
|
SYN
|
9/10
|
144
|
|
CHR
|
10/10
|
167
|
|
INSM1
|
10/10
|
206
|
|
Lung NET
|
SYN
|
3/3
|
58
|
|
CHRO
|
3/3
|
85
|
|
INSM1
|
3/3
|
170
|
|
Lung SMCC
|
SYN
|
48/51
|
103
|
|
CHR
|
43/51
|
80
|
|
INSM1
|
49/51
|
194
|
|
Lung LCNEC
|
SYN
|
1/1
|
80
|
|
CHR
|
0/1
|
0
|
|
INSM1
|
1/1
|
270
|
|
Nasal cavity
|
SYN
|
1/1
|
90
|
|
CHR
|
0/1
|
0
|
|
INSM1
|
1/1
|
90
|
|
BOT
|
SYN
|
1/1
|
80
|
|
CHR
|
1/1
|
120
|
|
INSM1
|
1/1
|
190
|
|
Presternal
|
SYN
|
1/1
|
185
|
|
CHR
|
1/1
|
250
|
|
INSM1
|
1/1
|
160
|
|
UB
|
SYN
|
1/1
|
200
|
|
CHR
|
1/1
|
190
|
|
INSM1
|
1/1
|
260
|
|
Cervix
|
SYN
|
5/5
|
163
|
|
CHR
|
4/5
|
74
|
|
INSM1
|
5/5
|
198
|
|
Vault
|
SYN
|
1/1
|
285
|
|
CHR
|
1/1
|
175
|
|
INSM1
|
1/1
|
300
|
Abbreviations: BOT, base of tongue; CgA, chromogranin A; GIT, gastrointestinal tract; INSM1, insulinoma-associated protein 1; LCNEC, large cell neuroendocrine carcinoma; NET, neuroendocrine tumor; SMCC, small cell carcinoma; SYN, synaptophysin; UB, urinary bladder.
Table 4
H-score comparison
|
Parameters
|
Number
|
p-Value[a]
|
|
SYN vs. INSM1
|
|
|
|
SYN < INSM1
SYN > INSM1
SYN = INSM1
|
79
|
0.001
|
|
18
|
|
3
|
|
CgA vs. INSM1
|
|
CgA < INSM1
CgA > INSM1
SYN = INSM1
|
76
|
0.001
|
|
21
|
|
3
|
|
SYN vs. CgA
|
|
|
|
SYN < CgA
SYN > CgA
SYN = CgA
|
53
|
0.195
|
|
46
|
|
1
|
Abbreviations: CgA, chromogranin A; INSM1, insulinoma-associated protein 1; SYN, synaptophysin.
a Wilcoxon rank test.
Discussion
NET is a relatively rare disorder and its diagnosis requires an integrated approach. Well-established neuroendocrine markers such as SYN, CgA, and CD56 are cytoplasmic markers and it is difficult to interpret them in small biopsy. In our study, we have evaluated the newly evolved marker INSM1 and compared it with various parameters which are statistically not significant. Our findings are concordant with McHugh et al.[13] In their study they have compared INSM1 in GEP-NENs. Total 97% cases were positive for INSM1, which was higher than the traditional NE markers. Correlation of expression of INSM1 with SYN and CgA showed statistically significant association in our study (p-value < 0.05). Rooper et al[14] studied INSM1 in all thoracic NETs and they have found statistically significant association of INSM1 with SYN and CgA (p-value < 0.001), which was concordant with our study. However, Zou et al[15] did not find statistically significant association of INSM1 with CgA and SYN (p-value < 0.09 and 0.494, respectively). In our study, sensitivity of INSM1, SYN, and CgA were 97, 96, and 86%, respectively. Comparison of sensitivity of INSM1 and conventional marker of previous study is discussed in [Table 5].
Table 5
Comparison of expression of NE markers with previous study
|
Studies
|
INSM1 (%)
|
SYN (%)
|
CgA (%)
|
|
Our study
|
97/100 (97)
|
96/100 (96)
|
86/100 (86)
|
|
Fujino et al[16]
|
100/102 (98)
|
88/102 (86.2)
|
84/102 (82.3)
|
|
Rosenbaum et al[8]
|
27/30 (90)
|
29/30 (96.7)
|
21/30 (70)
|
|
Aldera et al[17]
|
59/69 (85.5)
|
63/69 (91.3)
|
48/69 (69.5)
|
|
McHugh et al[13]
|
89/110 (82.9)
|
109/110 (99.1)
|
96/110 (87.3)
|
|
Mukhopadhyay et al[6]
|
144/152 (95)
|
147/150 (98)
|
125/149 (84)
|
|
Kriegsmann et al[18]
|
276/372 (74.2)
|
319/372 (85.8)
|
289/372 (77.7)
|
|
Rooper et al[14]
|
99/103 (96.1)
|
79/103 (76.7)
|
67/103 (65.0)
|
|
Sakakibara et al[19]
|
120/141 (85.1)
|
87/141 (61.7)
|
74/141 (52.5)
|
|
González et al[20]
|
32/32 (100)
|
32/32 (100)
|
31/32 (97)
|
Abbreviations: CgA, chromogranin A; INSM1, insulinoma-associated protein 1; NE, neuroendocrine; SYN, synaptophysin.
In our study, we evaluated INSM1 and other conventional NE markers by calculating H-scores. The mean H-scores of INSM1, SYN, and CgA in our study were 194.5, 119.6, and 108.7, respectively. This is slightly lower than the study by Fujino et al.[16] In their study, the mean H-scores were 211, 191, and 122, respectively. However, in our study comparison of INSM1 H-score with traditional NE markers was statistically significant. This result was concordant with the study of Fujino et al[16] (p-value < 0.0001).
Conclusion
INSM1 is a superior immunohistochemical marker when compared with traditional NE markers (SYN and CgA). Statistically significant association was found between expression of INSM1 and SYN and CgA. However, larger prospective studies should be undertaken to assess the site-based INSM1 expression as well as to investigate the role of INSM1 in prognosis of NEN.
